Luc wrote: "I told him recently several times: the frequencies in your RFM69 file don't work in rtldavis. They simply are all (very) different from the set I use. *** There is also no constant difference between the two frequency sets as he mentions. *** Note: The original US set in rtldavis is probably derived from the RFM69 set because the differences between these two sets are relatively constant. So far nobody could explain why these RFM69 frequencies don't work here."
First, unless I misunderstood something, the statements above marked with *** are contradicting each other. There are ONE and only ONE set of frequencies. The DO work for RFM69-based receivers. If the frequencies there don't work in rtldavis, the error is in the rtldavis code or in the receiver equipment, its offset, etc. The original author didn't verify his code using real transmitters? That's possible, but then we're dealing with a huge uncertainty from the get go. Please see the https://docs.google.com/spreadsheets/d/1SyNKF-ynoYvKEiOfr7CcZSvaKDU49rJLbNvfN0avzbQ/ sheet about the (almost) constant difference between rtldavis and the RFM69 frequencies (rtldavis xxx columns). I think the presence of this relatively constant offset is the offset error in someone's SDR stick (as things look, it might not even be the author's...) If you can, please try to do the visual test in SDR# as I mentioned in the other thread to see if your EU or US transmitters are in the ballpark at all or not. This is in my opinion very important, to at least eyeball a basic confirmation or mistake somewhere. Other notes. Around 90+% pctgod is very good for such a long time! I wouldn't even expect anything more from the SDR solution for a lot of reasons. What is the dimension of your freq. correction graph? Hz of kHz? If it's Hz, then it's basically spot on. However, that -1600 offset is horrendous. Something is going on, indeed. If you need to correct/scan the frequencies and such huge errors are present in reception, it must lie with something else, possibly in the demodulator code. I still stand by the statements about the correct frequencies in the RFM69 code. On Saturday, March 16, 2019 at 9:17:15 PM UTC+1, Luc Heijst wrote: > Hi Rich and Paul, > > Thanks for testing. > > When you look at my answer to kobuki, ALL IN CAPITALS, you must have > thought "qxqxwtr!" > > I know kobuki for a long time and he assisted me with the wind correction > formulas in the weewx-meteostick driver. > > I told him recently several times: the frequencies in your RFM69 file > don't work in rtldavis. They simply are all (very) different from the set I > use. > There is also no constant difference between the two frequency sets as he > mentions. > Note: The original US set in rtldavis is probably derived from the RFM69 > set because the differences between these two sets are relatively constant. > So far nobody could explain why these RFM69 frequencies don't work here. > > I'm used to do my own research and more than once the formulas of the > 'DeKays' and 'Kobukis' appeared not to be correct as I have learned with > the decoding of the raw Davis messages. > > The original rtldavis program had an automatic frequency correction per > channel mechanism build in and also a random start channel. > With these two features the program could 'find' more or less its own > frequency table. > > The problem is, that is only working for one transmitter. I tried to use > the mechanism for three transmitters simultanious and it did more harm than > good, so I have removed completely the correction mechanisme. > The problem the original author had, is that he didn't own any Davis > weather equipment and could not test the program himself. > > So, how now further? I think we are still on the right way with our > practical approach. > > Below the found hop frequencies so far. The yellow marked frequencies are > based upon Rich' test and the blue frequencies on Pauls test. > The white ones in between I have calculated on some assumptions and may be > totally wrong. > > The two graphs below show the pctGood percentage of the three stations > read by rtldavis and the 5-minute frequency errors of the 5 EU channels. > The rtldavis program was used with parameter -fc -1600 (new!) which is an > overal frequency correction for all (5 or 51) channels. > > > > The above graph show the pctGood percentage of the three stations read by > rtldavis. > rtld_total overal value > ISS-1 my first ISS > WIND my anemometer station included my solar sensor > LEAF-SOIL my leaf-soil station with 4 sensors (1 soil-temp, 2 > soil-moisture, 1 leaf-wetness sensor) > mstk-total the overal pct good value of the meteostik device which > measures the same equipment > > Cheers, Luc > > PS. I have uploaded the weewx-rtld driver which parses the rtldavis output > and stores the results in the weewx database. See: > https://github.com/lheijst/weewx-rtld > But first you have to get rtldavis working with the US frequencies before > you can test weewx-rtld! > > channel > newCorFreq > > 1 > 902355835 > > 2 > 902918911 > > 3 > 903360587 > > 4 > 903949606 > > 5 > 904372748 > > 6 > 904853902 > > 7 > 905359285 > > 8 > 905976213 > > 9 > 906386478 > > 10 > 906849336 > > 11 > 907358272 > > 12 > 907891789 > > 13 > 908438475 > > 14 > 908932739 > > 15 > 909468537 > > 16 > 909883984 > > 17 > 910434492 > > 18 > 910875526 > > 19 > 911370606 > > 20 > 911906039 > > 21 > 912369294 > > 22 > 912876933 > > 23 > 913412148 > > 24 > 913957365 > > 25 > 914453558 > > 26 > 914961673 > > 27 > 915409315 > > 28 > 915964902 > > 29 > 916417857 > > 30 > 916904836 > > 31 > 917399381 > > 32 > 917886807 > > 33 > 918396416 > > 34 > 918889783 > > 35 > 919399303 > > 36 > 919931359 > > 37 > 920479725 > > 38 > 920947548 > > 39 > 921471901 > > 40 > 921926728 > > 41 > 922719570 > > 42 > 922903934 > > 43 > 923947268 > > 44 > 923910143 > > 45 > 924413537 > > 46 > 924918736 > > 47 > 925452435 > > 48 > 925998868 > > 49 > 926424136 > > 50 > 927028377 > > 51 > 927230693 > > > ------------------------------ > [image: Avast logo] <https://www.avast.com/antivirus> > > This email has been checked for viruses by Avast antivirus software. > www.avast.com <https://www.avast.com/antivirus> > > <#2910243956.20190316171705@gmail.com_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> > -- You received this message because you are subscribed to the Google Groups "weewx-user" group. To unsubscribe from this group and stop receiving emails from it, send an email to weewx-user+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
